The first report of the mobile colistin resistance gene, mcr-10.1, in Kenya and a novel mutation in the phoQ gene (S244T) in a colistin-resistant Enterobacter cloacae clinical isolate

ABSTRACT This study describes the identification of the mcr-10.1 gene in a clinical isolate of an ST1 Enterobacter cloacae isolate cultured in 2015 in Kenya. The isolate was multidrug resistant, phenotypically non-susceptible to various antibiotics, including colistin. Whole genome sequence analyses indicated carriage of chromosomally encoded antimicrobial resistance genes and the colistin-resistant gene mcr-10.1 located on a 72-kb plasmid designated pECC011b with an IncFIA(HI1) replicon directly adjacent to tyrosine recombinase gene, xerC, and downstream of an ISKPn26 insertion sequence. Studies have shown that expression of mcr-10.1 may not be sufficient to confer colistin resistance, but a novel non-synonymous mutation (S244T) was identified in the phoQ gene known to influence colistin resistance within lipid modification pathways, which could have complemented the mcr-10.1 resistance mechanism. In silico analysis of the mutant phoQ protein shows the location of the mutation to be at the Histidine kinases, Adenyl cyclases, Methyl-accepting proteins and Phosphatases (HAMP) region, which plays a crucial role in the protein’s activity. This study and our previous report of mcr-8 in Klebsiella pneumoniae indicate the presence of mobile mcr genes in the Enterobacterales order of bacteria in Kenya. The study points to the importance of regulation of colistin in the animal industry and enhancing surveillance in both human and animal health to curb the spread of mcr genes and accurately assess the risks posed by these mobile genetic elements in both sectors. IMPORTANCE This paper reports the detection of new colistin resistance mechanisms in Kenya in a clinical isolate of Enterobacter cloacae in a patient with a healthcare-associated infection. The plasmid-mediated resistance gene, mcr-10.1, and a novel amino acid mutation S244T in the phoQ gene, located in a region of the protein involved in membrane cationic stability contributing to colistin resistance, were detected. Colistin is a critical last-line drug for multidrug-resistant (MDR) gram-negative human infections and is used for treatment and growth promotion in the animal industry. The emergence of the resistance mechanisms points to the potential overuse of colistin in the animal sector in Kenya, which enhances resistance, threatens the utility of colistin, and limits treatment options for MDR infections. This study highlights the need to enhance surveillance of colistin resistance across sectors and strengthen One Health policies that ensure antimicrobial stewardship and implementation of strategies to mitigate the spread of antibiotic resistance.

See the funding table on p. 5.
The material for this publication has been reviewed by the Walter Reed Army Institute of Research, and there is no objection to its publication.The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of the Army or the Department of Defense.The investigators have adhered to the policies for the protection of human subjects as prescribed in AR 70- 25. encapsulation, chromosomal mutations of lipid modification pathway genes, and the plasmid-borne mcr genes 10 variants (mcr 1 to 10) of these highly mobile genes have been detected in diverse sources globally, which could lead to the loss of this critical last-line drug (2,3).
In 2015, a colistin-resistant Enterobacter cloacae isolate was detected in a 23-yearold female inpatient at a hospital in Nairobi, Kenya, with a healthcare-associated leg wound infection enrolled in an antimicrobial resistance surveillance study (KEMRI2767/ WRAIR 2089).The bacterial isolate was identified using the GN-ID card and antimicrobial susceptibility test using the XN05 card on a VITEK2 (Biomerieux, Craponne, France).Colistin resistance was confirmed using a cation-adjusted Mueller Hinton broth dilution assay, including a Pseudomonas aeruginosa (ATCC 29853) negative control and a Proteus mirabilis positive control.Antibiotic resistance was interpreted according to the Clinical and Laboratory Standards Institute 2022 breakpoints (M100, 32nd edition) (4).Total DNA was extracted using the Quick-DNA Fungal/Bacterial Miniprep Kit (Zymo Research, California, USA).Nanopore sequencing libraries were prepared using the Ligation Sequencing Kit (LSK109) and loaded onto an R10.3 flow cell.Illumina sequencing libraries were prepared using the KAPA HyperPlus Kit (Roche Diagnostics, Indianapolis, IN, USA), quantified using the KAPA Library Quantification Kit-Illumina/Bio-Rad iCycler (Roche Diagnostics, Indianapolis, IN, USA), and sequenced with the MiSeq Reagent Kit v3 (600 cycles) on an Illumina MiSeq (Illumina Inc., San Diego, CA, USA).
The short reads were trimmed using Trimmomatic v0.39 (12) [to filter off low-quality reads (Q score <20) and reads less than 75 bp].The high-quality reads were then mapped against the reference Enterobacter cloacae ATCC 13047 genome (NC_014121.1)using bbmap v39.01 (13).The mapped reads were sorted and indexed with samtools v1.9 (14).Bcftools v1.9 (14) was used to generate genotype likelihood at each genomic position with coverage, followed by calling the variants using the default calling method with a minimum depth of 10.Low-quality variants (QUAL <40) were then filtered out.The filtered variants were annotated against a custom database built using the Enterobacter cloacae ATCC 13047 genome (NC_014121.1) with Snpeff v5.0 (15).Finally, the variants of pmrA/B, phoP/Q, and mgrB genes were identified in the output vcf files and visualized using IGV v2.16.0 (16).
The coding nucleotide sequences of wild type and mutant phoQ were translated using Expasy's Translate tool (17).Their 3D structures were predicted with Alpha fold2-multimer (18) and MARTINI coarse-grained molecular dynamics simulations were performed as described by Mahmood et al. (19) with slight modifications on the lipid bilayer composition to POPE:POPG (3:1).Visualizations were performed with Molviewer (20).
The hybrid assembly yielded a genome of 5,247,180 bp with five contigs, i.e., one chromosome of 5,065,027 bp, three plasmids of 99,848 bp, 72,193 bp, and 4,096 bp, and one unplaced contig of 6,016 bp.The GC content was 54.75% with a mean N50 of 5,065,027 (NCBI genome accession JARXZG000000000).The chromosomally encoded AMR genes confer resistance to antibiotics of the beta-lactams, fluoroquino lone, polypeptide, and fosfomycin classes and MDR efflux pumps (Table 1).The only plasmid-encoded gene was the mcr-10.1 colistin resistance gene.
The mcr-10.1 gene was on a plasmid designated pECC011b (Fig. 1).The plasmid had an IncFIA(HI1) replicon and was 99.5% identical to the pECC59-2 (CP080472.1)isolated from a clinical Enterobacter hormaechei in China.The mcr-10.1 gene was adjacent to xerC, a sitespecific tyrosine recombinase downstream of an ISKPn26 insertion sequence similar to previous reports of the genomic environment of mcr-10 (21,22).
Colistin resistance can also result from mutations in lipid modification pathway genes, including phoQ, phoP, pmrA, pmrB, and mgrB (2).A non-synonymous mutation (730T>A) in phoQ, leading to serine-to-threonine (S244T) change, was identified.A synonymous mutation (12G>A, L4L) was found in the phoP gene, while pmrA, pmrB, and mgrB genes had no mutations.To identify the location of the non-synonymous mutation, top-ranked structures for both wild type and mutant homodimer proteins were selected using Alphafold's ranking scheme.The predicted 3D structures of wild type and mutant phoQ exhibited a consistent arrangement, featuring a periplasmic sensor domain connected to a transmembrane domain, the Histidine kinases, Adenyl cyclases, Methyl-accepting proteins and Phosphatases (HAMP) domain where the S244T mutation was located, and the histidine kinase region.This structural consistency was maintained when the proteins were inserted into a lipid bilayer, as shown in snapshots during simulation (Fig. 2).
The mcr-10 gene was first described in China in an Enterobacter roggenkampii clinical strain (22).Since then, it has been identified in 18 countries on five continents, including Africa, and 15 species of seven genera in the Enterobacterales family (23,24).This study identified an E. cloacae isolate with an mcr-10.1 gene, the first report in Kenya and the only one globally in an ST1 isolate.The ability of the mcr-10-bearing plasmids to self-transmit, mobilize, and re-integrate indicates that the mcr-10.1 gene has a high potential to spread within the Enterobacterales family in Kenya.Future conjugation experiments would confirm the transmissibility of the mcr-10.1 gene in this context.The phoQ gene mutation was novel, so its potential impact on protein function was assessed in silico.Molecular dynamic simulations revealed that the S244T mutation was in the HAMP region.As demonstrated in a study by Mensa et al. (25), mutations in this region can disrupt the balance between the sensor and catalytic domains, potentially increasing kinase activity and causing cationic imbalance, ultimately affecting colistin susceptibility (25).
In summary, this study presents emerging mechanisms of colistin resistance in an E. cloacae isolate in Kenya, with the first report of the mobile mcr-10.1 gene, which fills gaps in mapping global antimicrobial resistance.Although this was detected in a single isolate, the mobility of the associated plasmid indicates a high risk of colistin resistance transmission within and between species.As colistin is considered a last-resort antibiotic for treating gram-negative bacterial infections and is used extensively in  animal production, robust policies are needed to enhance antimicrobial stewardship to curb mcr gene spread and preserve its efficacy.

FIG 1
FIG 1 The genetic organization of the plasmid pECC011b (NCBI accession JARXZG010000003) harboring the mcr-10.1 gene (highlighted in yellow) located downstream of ISKpn26 (highlighted in orange), and XerC (highlighted in cyan).The arrows indicate the direction of translation of the genes, and the circle represents the plasmid's size, 72,193 bp.Hypothetical proteins were hidden from the map.(The plasmid was annotated with Prokka and visualized with Snapgene.)

FIG 2
FIG 2 3D snapshots of the wild type phoQ (A) and mutant phoQ (B) obtained from trajectory files during simulation displaying the sensor domain, HAMP, and the catalytic domain.The protein is a homodimer (chain 1 in sea blue and chain 2 in brown).The amino acid residue at position 244 (highlighted in black) is serine in the wild type and threonine in the mutated protein.

TABLE 1
Detection of AMR genes, genomic location, and corresponding antibiotic class in the E. cloacae clinical isolate